-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hey Valenta,
the last week I also tried to setup lucene-search-2.0 and I discovered by analyzing the code, that there is another MW extension called OAI (http://svn.wikimedia.org/svnroot/mediawiki/trunk/extensions/OAI) which collects updates in a new table (see sql scripts). And there is an update daemon, which grabs this information and triggers the search daemon build in indexer. You can start the daemon by typing "java -Djdbc.drivers=com.mysql.jdbc.Driver -cp .:LuceneSearch.jar org.wikimedia.lsearch.oai.IncrementalUpdater -d <put in you wiki db here>". But there is one last thing I am fiddling with: I did not get the daemon to overwrite the old index with the updated index, instead of saving the updated index in another directory.
HTH, Bene
-----Original Message----- From: mediawiki-l-bounces@lists.wikimedia.org [mailto:mediawiki-l-bounces@lists.wikimedia.org] On Behalf Of Trey Valenta Sent: Friday, June 08, 2007 8:48 PM To: mediawiki-l@lists.wikimedia.org Subject: [Mediawiki-l] Questions on lucene-search-2.0
I was finally able to install the Lucene search engine using the code from http://svn.wikimedia.org/svnroot/mediawiki/trunk/lucene-search-2.0, but I have some questions regarding updating indexes. Options for the importer are listed as:
-n - create a new index (erase the old one if exists) -s - make index snapshot when finished -l limit_num - add at most limit_num articles -o optimize - true/false overrides optimization param from global settings -m mergeFactor - overrides param from global settings -b maxBufDocs - overrides param from global settings --snapshot <db> - make snapshot only for dbname
Can anyone suggest an ideal way for updating the indexes? I've setup an hourly cronjob to dump current wiki content, import the dump, and make an index snapshot when finished (-s). Is this preferred over just creating a new index each hour?
Thanks in advance, Trey Valenta
MediaWiki-l mailing list MediaWiki-l@lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/mediawiki-l